RESEARCH METHODS - people.tamu.edupeople.tamu.edu/~w-arthur/Winfred's stuff/Winfred, research...
Transcript of RESEARCH METHODS - people.tamu.edupeople.tamu.edu/~w-arthur/Winfred's stuff/Winfred, research...
Winfred, research methods, 04-23-10; rv 10-14-14 3
Research Methods• means of discovering truth• what is truth?
Winfred, research methods, 04-23-10; rv 10-14-14 4
Research Methods• means of discovering truth• what is truth?
– Riveda Sandhyavandanam "The is only one truth [but] people often see it in different ways"
– see also Ludwig Wittenstein (1889-1952) The metaphysics of space and motion and the wave structure of matter (WSM) gives absolute truth and meaning to language
Winfred, research methods, 04-23-10; rv 10-14-14 5
Research Methods• means of discovering truth• what is truth?
– VALIDITY?– a conclusion based on a piece of research
is valid when it corresponds to the actual or true state of the world
Winfred, research methods, 04-23-10; rv 10-14-14 6
Validity a key (maybe THE key) criterion in the
evaluation of any piece of research or test (measure) the appropriateness of inferences drawn
from data data = results of research study research
validity data = test scores test and measurement
validity
Winfred, research methods, 04-23-10; rv 10-14-14 7
Research Validity• two philosophies, world views, or approaches
to truth discovery– Donald Campbell and Donald Rubin
– Shadish (2010; Campbell and Rubin: A primer and comparison of their approaches to causal inference in field settings. Psychological Methods, 15, 3-17)
• Campbell's causal model methodologically and design driven
• Rubin's causal model quantitatively and statistically driven
Winfred, research methods, 04-23-10; rv 10-14-14 8
Research Validity• can we recreate or recover truth statistically?
– missing values and data imputation? [RMNET]– "we have no information as to why subjects’ fail to respond to
items; to assume we do is total unwarranted ignorance" (Romie Littrell, 04/17/10)
– meta-analysis?– Dieckmann, Malle, and Bodner (2009; An empirical assessment
of meta-analytic practice. Review of General Psychology, 13,101-115)
– control variables?– corrections for faking?
– "once test users take a wrong course, there is no going back to the choice point" (Cronbach, 1990, p. 521)
Winfred, research methods, 04-23-10; rv 10-14-14 9
Research Validity re Campbell’s causal model [CCM]• facets of research validity internal external statistical-conclusion construct
• threats to research validity• controls
Winfred, research methods, 04-23-10; rv 10-14-14 10
• internal validity—the validity of inferences about whether observed covariation b/n A (presumed treatment) and B (presumed outcome) reflects a causal relationship from A to B, as those variables were manipulated or measured
Threats● history ● attrition or mortality
● maturation ● selection
● testing ● regression
Controls?
Winfred, research methods, 04-23-10; rv 10-14-14 11
• history effects (events outside the lab) observed effect between IV and DV might
be due to events occurring between the pretest and posttest when these events are not the treatment of research interest
• maturation effects source of error in a study related to the
amount of time between measurements
Winfred, research methods, 04-23-10; rv 10-14-14 12
• testing effects effects due to the number of times
particular responses are measured resulting from familiarity with the
measurement instrument
• attrition or mortality effects the dropping out of some participants
before a study is completed, causing a threat to validity
Winfred, research methods, 04-23-10; rv 10-14-14 13
• selection effects result from biases associated with the
selection of, and assignment of research participants into groups
• regression effects tendency of participants with extreme
scores on first measure to score closer to the mean on a second testing a statistical threat
Winfred, research methods, 04-23-10; rv 10-14-14 14
• internal validity extent to which we can infer that a
relationship between two variables is causal or that absence of a relationship implies absence of cause is the observed effect real or artifactual? threats
• correction randomization
Winfred, research methods, 04-23-10; rv 10-14-14 15
• external validity—the validity of inferences about whether the cause-effect relationship holds over variations in persons, settings, time, treatment variables, and measurement variables
● population validity
● ecological validity
● temporal validity
Controls?Threats
Winfred, research methods, 04-23-10; rv 10-14-14 16
• other participants interaction of selection and treatment population validity
• other settings interaction of setting and treatment ecological validity
• other times interaction of history and treatment temporal validity
Winfred, research methods, 04-23-10; rv 10-14-14 17
• external validity generalizability
• enhanced or increased random sampling for representativeness
• trade-offs between internal and external validity?
Winfred, research methods, 04-23-10; rv 10-14-14 18
• statistical conclusion validity—the validity of inferences about the correlation (covariation) b/n treatment and outcome
● low statistical power
● violation of test assumptions
● poor test reliability
Controls?Threats
Winfred, research methods, 04-23-10; rv 10-14-14 19
• low statistical power power analysis
sample size (n) effect size (magnitude of effect) power (.80) alpha (p-value, .05)
• violations of statistical test assumptions• poor reliability of measures
• address threats adequate power, meet test assumptions, and use reliable measures
Winfred, research methods, 04-23-10; rv 10-14-14 20
• construct validity—the validity with which inferences can be made from the operations in a study to the theoretical constructs those operations are intended to represent
● loose connection between theory and study
● evaluation apprehension
● experimenter expectancies ("good-subject" response)
Controls?Threats
Winfred, research methods, 04-23-10; rv 10-14-14 21
• construct validity use of appropriate theories, theoretical
labels, or models to describe and explain phenomenon being studied
Fatal-crashes drivers
No-crashes drivers
Fatal driving crash involvement: Locus of control vs attribution theory
Locus of control measure
Fatal-crashes drivers = externals
No-crashes drivers = internals
T I M E
Winfred, research methods, 04-23-10; rv 10-14-14 22
• construct validity use of appropriate theories, theoretical
labels, or models to describe and explain phenomenon being studied
Fatal-crashes drivers
No-crashes drivers
Fatal driving crash involvement: Locus of control vs attribution theory
Locus of control measure
Fatal-crashes drivers = externals
No-crashes drivers = internals
All drivers
Locus of control
measure
Crashes assess relationship b/n LoC and crashes.
Do externals have more fatal crashes than internals?
T I M E
Winfred, research methods, 04-23-10; rv 10-14-14 23
threats• loose connection between theory and study• changes in participants' behavior as a result of
being studied "good subject" response Hawthorne effect social desirability evaluation apprehension responses to experimenter expectancies
• control or minimize threats double-blind procedures single-blind procedures deception
Winfred, research methods, 04-23-10; rv 10-14-14 24
• interrelatedness of different facets or dimensions of research validity
Winfred, research methods, 04-23-10; rv 10-14-14 25
• good research design or study– free from threats– no alternative explanations– permits robust conclusions about relationships between
study’s variables
• conceptualized [i.e., research] as the "discovery of truth", it then becomes blatantly obvious that no single [primary] study will be sufficient to permit a claim of truth discovery– at best, it can only be claimed to be another datum in the
cumulative body of knowledge that slowly moves us closer to the discovery truth
Winfred, research methods, 04-23-10; rv 10-14-1426
Philosophies of Causal Inference Conditions of Causality• contiguity• temporal precedence• constant conjunction
Winfred, research methods, 04-23-10; rv 10-14-14 27
A rooster crows each dawn and observes that shortly after (30 minutes or so), the sun rises. After nine (9) months of this, the rooster concludes that he is responsible for making the sun rise. The farmer disagrees with this assertion and in an attempt to disprove this claim, asks the rooster to crow at 10:00 pm one evening.
By doing this, which condition of causality is the farmer trying to assess?
Winfred, research methods, 04-23-10; rv 10-14-14 28
The Empirical Research CycleResearch process - summarized as 5-step sequence
Statement of the problem
Design of research study
Measurement of variables
Analysis of data
Conclusions from research
Winfred, research methods, 04-23-10; rv 10-14-14 29
Research Validity
• naturalness of the research setting– lab or field– "naturalness" or "artificiality" of the setting– field research typically employs a real-life
setting, while lab research is more contrived and artificial
Winfred, research methods, 04-23-10; rv 10-14-14 30
Research Validity
• degree of control– confounding and extraneous variables– manipulation—this is reflective of a high
degree of control– research designs that permit manipulation
are technically referred to as "experiments"
Winfred, research methods, 04-23-10; rv 10-14-14 31
• time longitudinal and cross-sectional designs• data collection protocols observational designs?• secondary research designs meta-analysis
Winfred, research methods, 04-23-10; rv 10-14-14 32
Experimental Designs
• random assignment and manipulation– pretest-posttest with a control group– solomon four-group– posttest only with a control group
– within- and between-subjects designs
Winfred, research methods, 04-23-10; rv 10-14-14 33
Quasi-Experimental designs
• self-selection groups• pre-existing groups
– nonequivalent control group designs– delayed control group– mixed factorials
– designs without control groups– interrupted time series– repeated treatment
Winfred, research methods, 04-23-10; rv 10-14-14 34
Correlational Designs
• predictive• concurrent• postdictive
Winfred, research methods, 04-23-10; rv 10-14-14 35
Survey Research
• measurement and assessment of opinions, attitudes, and such, usually by means of questionnaires and sampling methods
Winfred, research methods, 04-23-10; rv 10-14-14 36
Additional Design and Methods Issues
• time longitudinal and cross-sectional designs
• data collection protocols observational designs?
• secondary research designs meta-analysis
Winfred, research methods, 04-23-10; rv 10-14-14 37
Additional Design and Methods Issues
• control techniques and strategies– random assignment to groups– matching– within-subject designs
– order and carry-over effects– counter-balancing
– manipulation checks– instrumentation of response
– CMV
– building nuisance variables into the study [moderators]– statistical control– experimenter effects and bias reduction
– double-blind procedures– automation